OS config, suggest by DeekThink

To deploy a k3s cluster on Alpine Linux VMs, follow these steps:

1. Prepare Alpine Linux VMs

Setup VMs: Create Alpine Linux VMs (1 for the server, others as agents). Ensure they have:
- Minimum resources: 512MB RAM, 1 vCPU, 10GB disk (adjust based on workload).
- Network connectivity between nodes (ports 6443, 8472/UDP for Flannel, and 10250).

2. Configure cgroups on All Nodes

K3s requires cgroups for resource management. On each VM:

# Install necessary tools
apk add curl nano

# Enable cgroups
echo "default_kernel_opts=\"cgroup_enable=cpuset cgroup_memory=1 cgroup_enable=memory\"" >> /etc/update-extlinux.conf

# Update kernel and reboot
update-extlinux
reboot

3. Install k3s Server

On the server node:

# Install k3s server (with Flannel on eth0)
curl -sfL https://get.k3s.io | sh -s - server --flannel-iface=eth0

# Enable and start service (if not already running)
rc-update add k3s default
service k3s start

# Verify node status
k3s kubectl get nodes

Retrieve join token:

cat /var/lib/rancher/k3s/server/node-token

4. Join Worker Nodes

On each agent node:

# Replace with your server IP and token
export K3S_URL="https://<SERVER_IP>:6443"
export K3S_TOKEN="<NODE_TOKEN>"

# Install k3s agent
curl -sfL https://get.k3s.io | sh -

# Enable and start agent service
rc-update add k3s-agent default
service k3s-agent start

5. Verify Cluster

On the server node:

k3s kubectl get nodes

6. Access Cluster Remotely

Copy the kubeconfig from the server:

cat /etc/rancher/k3s/k3s.yaml

Replace 127.0.0.1 with the server’s IP in the copied k3s.yaml, then use it with kubectl.

Troubleshooting

Logs: Check logs with journalctl -u k3s (server) or journalctl -u k3s-agent (workers).
Firewall: Ensure ports 6443 (API), 8472/UDP (Flannel), and 10250 (kubelet) are open.
cgroups: Verify cgroups are enabled post-reboot with cat /proc/cmdline.

Optional: Use Specific Versions

Install a specific k3s version:

curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=v1.26.5+k3s1 sh -

This setup provides a lightweight Kubernetes cluster using k3s on Alpine Linux. Adjust network interfaces (--flannel-iface) or CNI plugins as needed for your environment.